Loudness War
The Loudness War ( English for "loudness war" ) refers to the tendency of the music industry , music in gradually higher loudness levels - not to be confused with sound strength levels - produce, produce an overall impression that sets itself apart from the other artists. On the one hand, this compression of the audio signal leads to a more constant "audibility" of the music, but also results in a high loss of dynamics ; which is why musicians repeatedly discuss whether the consequences of this development are really desirable.
Basics
This phenomenon can be seen in many areas of the music and advertising industries , particularly radio and music albums released on DVD and CD; in the latter case, the loudness war stems from an interest in producing CDs that either sound as loud as possible or are perceived as louder than CDs by solo artists or competing music labels .
The maximum volume ( amplitude ) of a CD is limited, and this limit is mostly reached in loud passages. Therefore it is not possible to increase the volume per se. Instead, the quieter passages are made louder. This only increases the subjectively perceived loudness while at the same time reducing the dynamic range . The result is compression and possibly even distortion . Sometimes the recording volume itself is increased, which leads to clipping (overdriving) in loud parts of the song.
In the advertising industry as well as in radio and television there are similar tendencies, for example to make one's own station appear louder than the competition, to set the block of advertising apart from the actual program by increasing the volume, or to make a single commercial stand out from others.
introduction
Loudness
Initially, a so-called loudness war broke out between VHF radio stations , a means of fighting a competition for ratings by listeners for the respective stations. Soon after, record labels began increasing the loudness of both their LP and CD productions. The main reason for this "arms race" lies in the (subjective) advantages of the louder version, namely a subjectively better hearing impression is created, which can be traced back to the way in which the human ear perceives sound pressure at different levels . In humans, the ability to react to different frequencies depends not least on the sound pressure level; consequently, the more the sound pressure level is increased, the greater the number of perceptible high and low frequency tones. Music recorded at higher levels can be more easily replicated in higher-noise environments, such as in a car, on a train, or on a busy shopping street.
Even with playback systems with poor sound quality such as B. Internet audio streams , medium wave radio , monaural television and telephones, higher levels are perceived as subjectively sounding better. Another reason for the loudness problem can be found with artists who are increasingly trying to match their loudness levels to those of the best-selling current CD productions.
criticism
In addition to audiophiles and hi-fi enthusiasts, these practices have also met with harsh criticism from some leading members of the music industry, including Grammy winner Doug Sax , sound engineer Geoff Emerick (known worldwide for his work with the Beatles from Revolver to Abbey Road ), and many other. Although Bob Dylan rejected this approach on the one hand strictly with the words . "Take these modern productions sounds - they're terrible, they only consist of sound, there is nothing more clearly defined, no voice - absolutely nothing" , are the CD editions of Dylan's recent albums “Love and Theft” , Modern Times and Together Through Life, on the other hand, are themselves relevant examples of heavy compression.
If a radio station plays a CD track, it is very likely that it will run through the station's own chain of signal processors ( e.g. Optimod ), which further reduces the dynamic range and brings it close to the level of the absolute amplitude - regardless of the actual loudness of the original CD production. Critics of the method were already calling for immediate changes in music production in terms of loudness level.
In August 2006, the A&R vice-president of the Sony Music Group label One Haven Music denounced the "Loudness Wars" in an open letter, alleging that sound engineers in mastering studios were either forced to make the audio material loud against their will or this volunteered from the outset to attract the interest of industrial officials. There are also petitions to get bands to re-publish their productions, with these re-releases being less distorted. This could already be a start to undermine the currently practiced worldwide procedure of increasing the loudness at the expense of the dynamic range or to end the "Loudness War".
Musical works whose key elements are loudness, as is the case with Ravel's Boléro , are sometimes still played, but practically willfully destroyed by the extreme dynamic compression.
Effects
The method of increasing the loudness of a CD so that it is louder than competing productions leads to distortion if used excessively .
In the digital environment, this is commonly referred to as " clipping ". Digital sound carriers can not display individual values ( samples ) that are higher than 0 dBFS . Thus, every time signal peaks exceed 0 dBFS, the resulting waveform has distortion. However, since some tones and sounds such as the "kick" of a bass drum in the drums only briefly reach their peak level and are also considerably louder than the rest of the signal, the sound engineer can make the recording louder by increasing the volume and accept that the peak levels of the "kicks" in the drums are actually "clipped". This is seldom registered by the average listener. However, if distortions occur too often in a piece of music or if certain signals sound clearly falsified, the listening experience will be uncomfortably affected. Most modern productions - mainly pop music , but also many classical and jazz CDs - have these clipping.
Since the signal in analog audio media "bends" as soon as it reaches the absolute maximum level of 0 dBFS, this can also be used in the digital environment, either by transferring analog audio material from tape (using tape saturation) to a digital recording medium or by using software that emulates this effect . In technical terms, this is sometimes referred to as "soft clipping".
This analogue distortion creates harmonics that are perceived by the listener as a small “crack” or “hiss”. The results of this effect are influenced by a number of factors: on the one hand, the characteristics of the original sound itself, and, on the other, the strength and type of distortion used. Since the distortion caused in the analog way is not "flattened" as with digital clipping, the results sound less unpleasant. The more overdrive effect is applied to the signal, the more distortion is generated, the strength of which is in a range from "barely perceptible" (very weak) to "very audible" and - just like with digital clipping - from different behavior of the different musical instruments when distorted. In the remaining cases, compression or limiting is mostly used . Although the resulting bias in the final mix ( "is Final Mix ") is minimized, but the expressiveness reduces the settling transients (eg the drums.) By this method significantly and can - once the settings are too aggressive - the Strongly modify the natural dynamic character of other instruments involved in the recording - in most cases with negative consequences for the sound image. Techniques for increasing loudness do not always influence so-called macrodynamics , that is, the relative volume differences in the various sections of the piece of music. For example, loudness-based signal compressors (which also include limiters ) only affect the “local” signal.
Mastering studios commonly use multiband compression to achieve a more homogeneous mix that is easier to level out - which also works better with cheap playback devices - or to achieve a very individual sound or artistic effect. "Radio-grade" compression, on the other hand, is applied to the music in order to fine-tune the volume levels in different sections of the piece of music, which makes the production more suitable for listening in the background or in noisy environments, but the dynamic expression of the entire piece is also moderate. to very badly affect.
In general, the distortion reduces the speech intelligibility, the differentiability of the melody decreases, instruments lose their sound character, and soft background melodies are "drowned out".
history
(Note: some of the examples below use root mean square (RMS ) values based on calculations of average CD audio samples , using the full scale of 0 dB as a reference value.)
The procedure for mastering, focusing on loudness, can at first glance be traced back to the beginnings of the CD, but on closer inspection it already existed at a time when vinyl records were still the standard recording medium. Many of the music labels put out samplers , and as soon as artists and producers found their song to be quieter than the other songs on the compilation, they insisted on having their song remastered so that it could "keep up" with the rest. Also, on many Motown releases, the limits on how loud a record could be were set higher and higher, and affiliate labels were well known for making "the loudest 45- RPM releases in the music industry."
Nevertheless, the vinyl format set physical limits for loudness and compression: too aggressive modification of one of the two properties or both resulted in the LP, maxi or single no longer being playable; which also provides an explanation for the fact that the increase in loudness could never reach the aggressive values of a digital production, whereby the latter was no longer subject to any limits. In addition, sound engineers in mastering studios now have much greater options than in earlier times: modern digital effects processors allow the engineer to have much greater control over the loudness of a song than ever before; z. B. You can use a brick wall limiter . In its digital form, this is able to process an audio signal in a predictive manner and thus limit the level without delay, while the analog variant usually does not work in advance, but can only react to the signal, which results in a minimal reaction time of a few Milliseconds.
The stages of development of the increasing loudness increase in the CD can be roughly distributed over the first two and a half decades since the CD medium came into existence.
Since the CD was not the standard medium in pop music until the end of the 1980s, there was little motivation to increase loudness in order to bring the pieces in line with the loudness of competing products. In addition, CD players were very expensive and therefore generally reserved for high-end systems, on which the higher recording levels used during this time had a rather negative effect. In addition, for the first four or five years after the introduction of the CD, there was no way of compressing a recording in a digital environment, as digital signal processors were not yet available in the mastering studios at the time.
In summary, it was at two common methods of CD mastering either about adjusting the highest level of production at 0 dB full or nearly full scale, or the digital level in relation to the indicated levels of something more familiar Vu-meters to set . When using VU meters, a certain level (usually −6 dB, or 50% of CD amplitude on a linear scale) was set, which acted exactly like a saturation level (defined as 0 dB) of an analog production, adding a few dB of des CD recording level for potential swings in the so-called "red zone" (shown by a red hatched area in the display of the VU meter) - because digital recording media cannot exceed 0 dB in level. At that time, the RMS level of a randomly picked rock song was around −18 dB.
By the end of the 1980s, productions that exceeded this level came onto the market, and it already looked at that time that CD volume levels were increasingly likely to break through the digital limit; provided that the signal amplification would clipping no more than about two to four digital samples . This led to productions where the VU meter hit close to 0 dB on a randomly picked pop CD with a powerful beat or a rock CD and rarely reached the actual 0 dB. The album Appetite for Destruction by Guns n 'Roses provides an example from the early years, which have all the tracks, on average, an RMS level of -15 dB.
In the early 1990s, some sound engineers employed in mastering studios decided to take this method one step further and to use the CD levels in the same way as those of an analog tape and equate the digital full scale of 0 dB with the analog saturation level, so the production was so loud that every (or nearly every) beat hit 0 dB or more. Although some early albums from this period exist (eg. As the 1991 appeared, unofficially known as "Black Album" from Metallica ), virtually no album came before 1992 on the market, which had been mastered in this way. Two examples from 1992 are Dirt by Alice in Chains and Angel Dust by Faith No More . During the entire decade, the loudness of CDs varied in extreme ways, depending on the respective views of the sound engineer and other people involved in the mastering studio.
In 1994 this process called “ hot mastering ” was used everywhere, although there were exceptions: the album Superunknown by Soundgarden is one of these exceptions, as it was still mastered in the “old style”. The average loudness of a rock CD, expressed in RMS, was −12 dB.
By and large, most rock and pop CDs released in the 1990s were produced using this process.
The interest in releasing “ hot mastered ” CDs gradually met with approval from the music industry officials, not only because customers preferred to listen to the louder CDs. Sound engineers, musicians and labels independently developed their own concepts for making CDs louder. While the loudness of CDs was increased bit by bit in the 90s, some planned to get the most out of this process: so the majority of the pieces on (What's the Story) exceed Morning Glory? from Oasis already the -8 dB, which was very rare back then (1995).
In 1997 Iggy Pop assisted his band The Stooges with the remastering and remixing of the 1973 album Raw Power , which produced an album that was unique in every way: the loudest rock CD ever recorded, with an RMS of −4 dB, which occurs very rarely even for today's standards.
Another extreme example is the Japanese noise rock band Guitar Wolf , who consider the acoustic distortion of signal overdrive as a legitimate design element for their aesthetics and record their CDs far beyond the clipping limit.
After the turn of the millennium, loudness standards would eventually peak.
Regardless of the last few years, in which −10 dB was still the quasi-standard, it is currently being raised to −9 dB or even higher. Exceptions to this rule are very difficult to find these days. An often cited example of extreme loudness increase is the 2001 CD Their Greatest Hits: The Record by the Bee Gees . Universal Records had all of the tracks remastered, with many of them controlling to an average of −10 dB, which in some of the older songs led to distortions on the relatively sonorous bass track.
To extensive controversy in 2008 released album topped Death Magnetic by Metallica , the analysis under constant different, according to critics clipping suffer.
Many of the Toten Hosen's recordings are also mastered so loud that they contain audible distortion. Well-known examples are the refrains of “10 little Jägermeister”, “Days like this” and “Why am I not full?”.
Opinions
Opinions about the impact of the “ Loudness Wars ” are almost exclusively subjective. Proponents of “louder” CDs believe that consumers would prefer louder CDs anyway and that they are also a better choice for noisy environments.
Another group, in turn, believes that in the discussion of only a small number of albums should be considered (eg the 1999's Album. Californication of the Red Hot Chili Peppers : an album that occurred in the high-frequency digital clipping in such great numbers that hi-fi enthusiasts filed many complaints), while still others believe that any CD that is often based on the digital full scale should be classified as "unacceptable". In any case, many music listeners do not really notice the difference without a direct comparison with the original. Some think it is just a lesser evil. Yet others feel unable to listen to an album that has been mastered using a loudness-based method from start to finish.
In the September 2011 issue of Sound on Sound , a leading magazine in music technology, Emmanuel Deruty published an article in which he vehemently denied that the “Loudness War” reduced the variability of dynamics in modern music. An indication of this is the fact that compared to an analogue template, the digital template of modern recordings has much higher dynamics. This argument is taken up and further elaborated in the article by Deruty and Tardieu, which appeared in the Journal of the Audio Engineering Society in January 2014.
The starting point for Deruty's and Tardieu's criticism was Suhas Sreedhar's article, The Future of Music , published by IEEE Spectrum in 2007 . They found that Sreedhar initially interprets the crest factor of a song as the difference (in decibels) between the maximum level and the average level, which he then calls the dynamic range . Then Sreedhar would go on to explain how reducing this dynamic range leads to a constant sound level : "[It] would be as if someone was shouting out everything he or she said". He is now referring to the variability of the volume. The same argument was later repeated several times in popular non-academic publications such as The Wall Street Journal .
According to Deruty and Tardieu, such a sloppily executed generalization of dynamics would only stir up confusion between the terminology of the crest factor (peak value to RMS) and dynamics in the musical sense (pianissimo to fortissimo). It would also contribute to a lack of understanding of the causes of the volume war. They suggest that the term dynamic range should be avoided altogether for lack of a proper definition.
Deruty and Tardieu then evaluated the evolution of both the crest factor and the R128 Loudness Range (LRA) over a 45-year period. They find that the LRA remained stable at all times, although the crest factor decreased significantly between 1985 and 2010. Thus they conclude that the macrodynamic variability was not significantly affected by the volume war. This conclusion was supported six months later by the publication of Joan Serrà in Nature's Scientific Reports .
The form of the evaluation was in turn questioned by others. It has been suggested that the LRA (which Deruty used to evaluate variability) was designed to measure volume variation within a track (sometimes called macrodynamics ), while the R128 Peak-to-Loudness Ratio (PLR) is a measure of the microdynamics is - similar to the crest factor. Shepherd and Katz, e.g. B., claim that the PLR is a better measure of assessing the total perceived dynamic range. And this clearly shows a downward trend in the course of the 1990s. TC Electronic's senior test engineer Esben Skovenborg takes the position that the crest factor does not contain a sufficient description for the dynamics of music. Why this is the case and why the LRA can be considered more suitable, he explained in his contribution to the AES Convention in 2012.
As of today (October 2014) there is no agreement as to whether the dynamic range should be expressed as LRA or PLR. In their 2010 contribution, Jon Boley et al. it is not possible to establish a direct connection, neither between the LRA nor a variant of the PLR and the perception of the musical dynamics. In addition, it should not be overlooked that the general public is unlikely to have sufficient sensitivity to dynamic compression, as Jens Hjortkjær and Mads Walther-Hansen point out in their 2014 article in the Journal of the Audio Engineering Society . The authors investigated whether dynamic compression has a negative effect on sound quality. They note that “the results provide no evidence of the effects of dynamic range compression, either on personal preference or perceived depth. The data suggest that the listeners are less sensitive than generally assumed, even with a high degree of compression. "
Other formats
Other audiovisual media formats
At the moment , the Loudness War seems to mainly affect audio CDs and - consequently - all reproductions in digital form (MP3, etc.).
In some cases, however, newer releases and remasters on vinyl records are also affected, as special mastering is necessary for this format; however, these masters are often created directly from the digital masters. The technical limits of the "analog" material make it necessary to reduce the volume before pressing so that the high loudness has no effect on the end result. When looking at the wave form of such vinyl recordings (see picture on the right) the impression of a more dynamic master can be wrongly created, this is due to the distortion of the (electro) mechanical production and playback process. But there are also examples of recent releases where the vinyl version actually stands out positively from the digital version.
Some releases on SACD and DVD- Audio are also affected by the method; Regardless of this, almost all DVD-Audio sound carriers contain either a Dolby Digital (AC3) or DTS audio track so that the DVD can also be played on a standard DVD player that does not support "real" DVD-Audio playback. Dolby Digital has a well-balanced, well-defined playback level, which the DTS audio track will likely adopt over time. It is therefore advantageous that the high-resolution audio track is based on the same reference level during production - which is usually the case.
Since the target group of these new, high-resolution sound carrier formats consists largely of audiophile music lovers, any attempt to apply loudness-oriented mastering to the audio tracks would be highly counterproductive, as the targeted group is known to be very critical in terms of sound quality and dynamics.
Problem of the commercials
Commercials in Germany have been increasingly affected by the loudness war since the early 1980s . Spots are generally compressed more and more in order to make the individual spot sound louder than other spots and the commercials as a whole louder than the rest of the program. The commercial is intended to take radio and television participants by surprise with the advertising message. The pleasant loudness curve of the transmitter inevitably falls by the wayside. Radio and television participants are therefore expected to make leaps in loudness that were generally considered unreasonable a few years earlier.
This effect has been steadily and significantly strengthened for years, because the control guidelines of the transmitters have also been significantly changed. The control guidelines still applied at ZDF in 1984 provided, for example, to control advertising blocks 9 dB less than voice broadcasts. Ten years later it was already common practice for many private broadcasters to control advertising 9 dB more than program contributions. Almost all commercials are already available in a much more compressed form than program contributions. In parallel, a very similar development took place in radio.
Various workarounds for the volume problem, which an increasing number of viewers find annoying, have now been developed for television. B. external additional devices that limit the maximum loudness of the looped signal to a set value or the receiving devices themselves offer a function to reduce the volume during advertising by pressing a button by a previously defined value via remote control.
Possible ways out and solutions
- From August 31, 2012, the public and private television programs in Germany planned to control their audio levels in accordance with the new R-128 control directive of the European Broadcasting Union (EBU). For this purpose, a new measurement method and the new Loudness Units Full Scale ("LUFS") were developed, which should make it possible to control the sound level according to its "loudness" and no longer according to its peak level. All workplaces where sound processing takes place should be equipped with new types of measuring instruments. As a result, highly compressed commercials would no longer sound louder than news or movies, for example. There would also no longer be any sharp jumps in volume between different television programs. In practice, however, nothing has happened in this regard; Even today, TV advertising still sounds louder than the rest of the program almost everywhere.
This method of adjusting the loudness is also suitable for radio programs, but there is still no binding introduction date for radio programs.
- Relocation of the compression process to the end device of the music listener by providing support for compression and limiter settings for consumer hi-fi components. Whether and how much compression is used is then up to the listeners. (The established " loudness " buttons on amplifiers have nothing to do with compression or limiting, so the term "loudness" has nothing to do with the loudness war discussed in the article . Rather, the "loudness" function is a sound influencing function the position of the volume control.)
- In 2001 a new standard for digital audio formats was proposed: Replay Gain . Software that can read out the replay gain calculates the loudness of an individual track or album as perceived by the ear and stores the level values in the file's tags so that software that can play the file and support replay gain can record the playback level can set a standard level. When levels are adjusted this way, heavily compressed tracks can often sound a little quieter than older audio material that has not been compressed as much. The most effective evidence can be provided by comparing an original CD release with its later remaster.
This process known as " volume leveling " is also supported by some well-known players, e.g. B. "Sound Check" in iTunes or "Volume Leveling" in Windows Media Player . It should be noted, however, that " volume leveling " - which also includes that used by Replay Gain - can only reduce the volume of the loudness-based mastered audio material, but cannot recalculate "lost" dynamic values due to excessive compression, nor avoid the consequences of clipping. - The DualDisc is a double-sided CD with the album on one side in CD-Audio (usually mastered for maximum loudness) and on the other in DVD-Audio using "milder" mastering. This creates a sensible compromise: the page with “CD-Audio” can be played in the car or in the mp3 player and thus show off its advantages of strong, loudness-oriented compression; the other side can be played at home on hi-fi equipment so that the music listener can enjoy the full dynamic range of CD production.
- HDCD (High Definition Compatible Digital) is a technology that uses dynamic range compression when mastering a CD. If the HDCD is played on a standard CD player, the sound is compressed and appears “loud”. However, if the HDCD is played on a special CD player that supports HDCD, the module for expanding the dynamic range comes into play, which reverses the compression values used during mastering into their opposite. If the acceptance of HDCD were better, it would be possible for the average consumer (who likes to listen to his music in the car or on the mp3 player) to listen to his music “loud” with a lot of compression, while the other consumer uses his HDCD player to listen to the music can enjoy using the full dynamic range. Unfortunately, however, very few CD players have this option for decoding HDCD data, so this currently still represents a tiny niche market; on the other hand, licensing issues need to be clarified when the technology is used: For example, the decoded signal must not be made available for digital output and thus “tapped on the way”.
literature
- Siegfried Wirsum: Practical sound technology, device concepts, installation, optimization. Franzis, Munich 1991, ISBN 3-7723-5862-4 .
- Rolf Beckmann: Manual of PA technology. Basic component practice. 2nd Edition. Elektor, Aachen 1990, ISBN 3-921608-66-X .
Web links
- Niels Boeing and Jochen Reinecke: “Loudness War - Volle Dröhnung”, zeit.de, March 4, 2012
- Marko Pauli: “The power of music - pop music in the loudness madness”, SWR2 Wissen, December 23, 2008
- Website of the “Turn Me Up! - Bringing Dynamics Back To Music " (English)
- dr.loudness-war.info: "Dynamic Range Database" for artists and albums
- Pleasurize Music Foundation on Loudness War German Podcast
Individual evidence
- ↑ Joe Gross: Everything Louder Than Everything Else. October 2, 2006, archived from the original on August 19, 2012 ; Retrieved May 20, 2013 .
- ↑ Sean Curnyn, "Tears of Rage: The Great Bob Dylan Audio Scandal," September 3, 2009
- ↑ NME Album Reviews - Jet Generation , http://www.nme.com/reviews/guitar-wolf/1216, accessed February 18, 2014
- ↑ a b Emmanuel Deruty: 'Dynamic Range' & The Loudness War (en) . In: Sound on Sound , September 2011. Retrieved October 24, 2013.
- ↑ Emmanuel Deruty, Damien Tardieu: About Dynamic Processing in mainstream music . Journal of the Audio Engineering Society. January 2014. Retrieved June 6, 2014.
- ↑ The Future of Music: Tearing Down the Wall of Noise ( Memento of the original from October 14, 2007 in the Internet Archive ) Info: The archive link was inserted automatically and has not yet been checked. Please check the original and archive link according to the instructions and then remove this notice.
- ^ Measuring the Evolution of Contemporary Western Popular Music . In: Scientific Reports . 2, July 26, 2012. doi : 10.1038 / srep00521 . Retrieved July 26, 2012.
- ^ Ian Shepherd: Why the Loudness War hasn't reduced 'Loudness Range' . Retrieved February 6, 2014.
- ↑ Jason Victor Serinus: Winning the Loudness Wars . Stereophiles. Retrieved February 6, 2014.
- ↑ Esben Skovenborg: Loudness Range (LRA) - Design and Evaluation . AES 132nd Convention. April 2012. Retrieved October 25, 2014.
- ^ Jon Boley, Michael Lester and Christopher Danner: Measuring Dynamics: Comparing and Contrasting Algorithms for the Computation of Dynamic Range . AES 129th Convention. November 2010. Retrieved October 25, 2014.
- ↑ Jens Hjortkjaer, Mads Walther Hansen: Perceptual Effects of Dynamic Range Compression in Popular Music Recordings . Journal of the Audio Engineering Society. January 2014. Retrieved February 18, 2015.
- ^ Ian Shepherd: Why the TT Meter doesn't work on vinyl (English). Retrieved November 4, 2019.
- ↑ EBU guideline R 128 in German (PDF; 401 kB)
- ↑ Explanations of the EBU on guideline R 128
- ↑ New method for loudness measurement (PDF; 1.1 MB)
